Dissertation > Excellent graduate degree dissertation topics show

The Application of Support Vector Machine in Neuron Classification with Spatial Structure

Author: ZengJie
Tutor: HanFengQing
School: Chongqing Jiaotong University
Course: Applied Computer Technology
Keywords: Fractal geometry Unbalanced data The signal to noise ratio indicators Support Vector Machine Neural Network Classifier
CLC: Q42
Type: Master's thesis
Year: 2012
Downloads: 12
Quote: 0
Read: Download Dissertation

Abstract


The brain is the human’s organization which has the most complicated structure andfunction and it contains kinds of neurons. In order to study the human brain the neuronmust be studied which comes from the brain, and in order to know neuron it isnecessary to know what the neuron is and how the neuron is classified. It is a veryimportant task that quickly and accurately classifying neuron. This task can makepeople easy to know the neuron. It is very meaningful and worthwhile to do this task.The main research of this paper is neuron and the entrance of research is the spatialstructure of neuron. It is classified by the feature of the spatial structure. It needs tocombine kinds of methods in the spatial structure of neuron. At last the neuron will beclassified by Support Vector machine.The main innovations are as follows:(1) The spatial structure of the neuron has the feature of fractal geometry in somewhere.By observation and comparison the spatial structure and to use the feature of its own,it is proposed an algorithm of fractal dimension and the fractal dimension of neuronis calculated. The fractal dimension is proved to classify the neuron is a valid anduseful by experiment.(2) The data of spatial structure of neuron is from neuromorpho. It is a database whichcollects the data of various biological laboratories. Because of the data is fromvarious laboratories, the data forms the unbalanced problem. In this paper, itcombines the common way of solution the unbalanced problem and proposes afitness to solve the unbalanced problem. And then it is proved is effective byexperiment.(3) The feature selection of spatial structure of neuron. There are many features fromthe spatial structure. It is a very important problem that how to choose the bestfeather of spatial structure. In this paper, it uses the signal to noise ratio to solve thisproblem. The signal to noise ratio (SNR) is easy to understand and calculate.To useSNR can get the better feather group of spatial structure.And then there will be aexperiment to prove it is valid.(4) After the data pre-processing over. At last, it will use the SVM to classify the neuron.By chose different kernel function, it gets the best result of classifying the neurons. Then there will be an experiment to prove it.Finally, the highest classification result is86.7%.It is from the way which combinesthe various methods. And the result shows that this method is valid to classify theneuron. At the same time it gives a new way to classify the neuron.

Related Dissertations

  1. Research of Diagnosing Cucumber Diseases Based on Hyperspectral Imaging,S436.421
  2. A Long-term Forecasting Model for Chaotic Time Series,O415.5
  3. Soft-sensing Technology in the Ethylene Distillation Process Applied Research,TQ221.211
  4. The Analysis of Taste in Cigarette Mainstream Smoke and Establishment of Forecasting Model,TS411
  5. Research and Realization of License Plate Character Recognition Algorithm Based on SVM,TP391.41
  6. Assessment of Antioxidant Level of Tomato Based on FT-IR Technology,O657.33
  7. Study on Fast Detection of Gray Mold on Plants Using Spectral and Multi-Source Spectrum Imaging Technology,S432.4
  8. Wake Field Prediction and Optimization of Marine Propeller Considering Cavitation Performance,U664.33
  9. Research of Preprocess Technique for 3D Space Handwriting Based on 3D Accelerometer,TP212
  10. A Study of Marine Oil-spill Identification System Based on Information Fusion,X55
  11. Prediction of Protein-ligand Interactions Based on Chemical Preference & Construction of Human Cytochrome P450 Substrate Database(CYP-Meta),R91
  12. A Study on Ship Structural Rational-based Design Method Based on Superelement and Support Vector Machine,U662
  13. Fault Diagnosis on Fuel Control System of a Certain Aero-engine,V263.6
  14. Based on support vector machine textile and apparel Safety Risk Assessment Model,TS941.1
  15. SVM based on ant colony clustering algorithm in intrusion detection,TP393.08
  16. Intelligent robust AQM algorithm and its equity research,TP393.06
  17. The Classification of Hyperspectral Remote Sensing Image Based on Hyper Sphere Multi-Class Support Vector Machine,P237
  18. Research Remote Image Classification Based on Support Vector Machine,P237
  19. Research and application of time series of the causal relationship between Grainger triggered the complex network analysis method based on,O157.5
  20. Classification Algorithm Based on Logistic Regression and Support Vector Machine,TP301.6
  21. Short-term Urban Traffic Forecasting Based on Multi-kernel SVM Model,U491.14

CLC: > Biological Sciences > Physiology > Neurophysiology
© 2012 www.DissertationTopic.Net  Mobile